Errors in Positive cell detection in large tissue images


We are analyzing using Positive cell detection (version 0.2.2) positive and negative nuclei of gH2AX labeled nuclei. We created a project and a script.

when getting the results some of the tissues appear to present different detection depending on the “squares” Qupath is using for analizing the entire section. That differ in some areas when repeating staining.

How can we solve that?

You can fins all data in the following link:

Many thanks in advance!

Here images!

I’m afraid that’s a known and expected limitation of QuPath’s cell detection, and the need to break up the region into small tiles. See here for an explanation and some suggested workarounds:

1 Like